[A] Relevance Theory
Total Page:16
File Type:pdf, Size:1020Kb
Wilson, Deirdre Nicholas Allott Wilson, Deirdre Nicholas Allott University of Oslo [email protected] Word Count: 2542 Reference Word Count: 422 This is a postprint (author’s final draft, after refereeing). The paper was published as: Allott, N. (2014). Wilson, Deirdre and Relevance Theory. In C. A. Chapelle, K. Aijmer, M. Gonzales- Lloret, L. Ortega, L. Plakans & B. Wolter (Eds.), The Encyclopedia of Applied Linguistics: Online Version. Wiley-Blackwell. Please send any comments or questions you have to: [email protected] Deirdre Wilson (born 1941) is one of the key figures in the history of pragmatics since its emergence as a distinct field of research in the 1970s. She is professor of linguistics (now emerita) at the Division of Psychology and Language Sciences at University College London, and, since 2007, a Research Director of the Centre for the Study of Mind in Nature at the University of Oslo. With Dan Sperber she is the co-founder of relevance theory, a theory of cognition and communication that is one of the leading programmes of research in linguistic pragmatics and related areas. She has worked on many topics in pragmatics and philosophy of language and mind, with Sperber, with other co-authors, and separately, including bridging, epistemic vigilance, Grice’s theory of conversation, inference, irony, literary interpretation and poetic effects, loose talk, metaphor, mind-reading/Theory of Mind, modularity of mind, mutual knowledge, non-declarative sentences, presuppositions, rhetoric, the role of prosody in pragmatics, the semantics-pragmatics distinction, and word meanings and lexical pragmatics. Wilson’s first degree was in Philosophy, Politics and Economics, followed by a BPhil in Philosophy and a year as a Lecturer in Philosophy in Oxford. Her early work in linguistics, including her PhD thesis, supervised by Noam Chomsky at the Massachusetts Institute of Technology, was on semantic and pragmatic presuppositions, and was influenced not only by Chomsky but also by the philosopher Paul Grice, whose lectures on ‘Logic and Conversation’ are generally seen as laying the foundations of modern pragmatics. Most of her research has involved relevance theory: developing it with Dan Sperber; using it as a framework for the study of communication; and revising and extending the theory. Several of the many students she has supervised have become leading researchers in linguistics, psychology or philosophy, including Diane Blakemore, Richard Breheny, Robyn Carston, Billy Clark, Ernst-August Gutt, Elly Ifantidou, Anna Papafragou, Stephen Neale, Tomoko Matsui and Tim Wharton. Wilson has published more than fifty papers, and five books: a monograph and a collection of papers written with Dan Sperber (Sperber & Wilson, 1986; Wilson & Sperber, 2012b) a monograph with Neil Smith (Smith & Wilson, 1979); a monograph (Wilson, 1975) based on her PhD thesis; and a novel, Slave of the Passions. In what follows, the essentials of relevance theory are explained, and two important strands of Wilson’s work are described: her theory of irony, developed with Dan Sperber, and her research with Sperber and with Robyn Carston and others on lexical pragmatics. Readers should also consult the entry on Dan Sperber in this volume for further coverage of their joint work. [A] Relevance theory 1 Wilson, Deirdre Nicholas Allott Relevance theory is an attempt to provide a psychologically realistic account of cognition and communication. It postulates that human cognition tends to allocate attention and processing resources so as to achieve the greatest cognitive effect for the least effort, that is, to maximise relevance, where relevance is defined as a positive function of cognitive effects and a negative function of processing effort. This is the Cognitive Principle of Relevance. A specific consequence of this general tendency is described in the Communicative Principle of Relevance. When a speaker makes an utterance, she takes up some of the hearer’s attention, so she communicates a (fallible) assurance that the utterance will be worth processing: that it will provide at least enough cognitive effects to repay the effort expended, that is, it will be at least relevant enough to be worth the hearer’s attention. In fact, Sperber and Wilson argue that hearers can expect speakers to go beyond this minimal level of relevance whenever they can see a way of doing so without going against their own interests or preferences, since this makes their utterances more likely to be understood. They call this the presumption of optimal relevance. The Communicative Principle of Relevance is that each utterance communicates such a presumption. The fact that utterances raise expectations of relevance, they argue, justifies hearers in applying a particular strategy in comprehension, the relevance theoretic comprehension procedure (or heuristic): to search for an interpretation by following a least effort path – i.e. to generate and assess interpretations starting with the one that is most accessible – and to accept the first interpretation that satisfies their expectations of relevance. The justification for the least effort path is that if the interpretation lies off it, then the hearer has been put to effort that he might have been spared, so the utterance, on that interpretation, will be less relevant than the speaker could have made it. Similarly, the hearer is entitled to accept the first interpretation he arrives at that satisfies his expectations of relevance, because searching for another would require more effort and the utterance, on that interpretation, would again be less relevant than the speaker could have made it. Relevance theory is inspired to a considerable degree by Paul Grice’s work on conversation and meaning, although it offers a very different account of communication from his. Many of the differences are due to relevance theory’s commitment to psychological realism, with a corresponding shift away from Grice’s concern with conceptual analysis. For example, Grice’s well-known definition of speaker (or ‘non-natural’) meaning in terms of three nested intentions is the inspiration for relevance theory’s claim that there are two speaker intentions central to communication, the informative intention and the communicative intention. The informative intention: The intention to inform an audience of something. The communicative intention: The intention to inform the audience of one’s informative intention. (Wilson & Sperber, 2004, p. 611. Cf. Sperber & Wilson, 1986, pp. 46–64.) Actions made with both types of intention are ostensive stimuli: utterances that speakers make with the open aim of imparting information. According to relevance theory, these are a psychological natural class defined by the informative and communicative intentions, and humans have a cognitive capacity which is dedicated to their interpretation. This is an empirical claim (in contrast with Grice’s definition of the concept of meaning). A second key postulate of relevance theory, that ostensive communication is essentially inferential, is also inspired by Grice. He pointed out that speakers can and often do convey much more than they explicitly say: they can also intentionally imply, or implicate. Relevance theory takes this insight further, proposing that both what the speaker intends to imply and the proposition(s) that she intends to express must be inferred. Linguistic decoding is part of the input to this process when what is 2 Wilson, Deirdre Nicholas Allott uttered is a linguistic phrase, but it is not essential to the process, since the same inferential ability is used in interpreting non-linguistic gestures. From Chomsky’s work on grammar, relevance theory adopts the criterion that a good theory of a human cognitive ability should be explicit in the sense that it should explain rather than presuppose our abilities (Sperber & Wilson, 1986, p. 94). Sperber and Wilson also take psychological tractability of the processes postulated to be crucial since their theory of utterance interpretation is an account at the level of performance rather than competence. Other central claims that cannot be covered here due to limited space include three distinctions original to relevance theory: i) the distinction between two types of meaning that words may encode: conceptual and procedural (Blakemore, 1987; Wilson & Sperber, 1993; Wilson, 2011). ii) the distinction between strong and weak communication, and the related claim that communication is aimed at increasing the manifestness of information to the hearer (rather than at getting the hearer to know it, or to mentally entertain it) (Sperber & Wilson, 1986, pp. 39–46, 59–60, 196–200; Sperber & Wilson, 2008, §§7–8). iii) the distinction between descriptive and interpretive use (Sperber & Wilson, 1986, p. 224ff.), which is the foundation of Wilson and Sperber’s analysis of verbal irony (discussed below) and interrogatives (Wilson & Sperber, 1993). [A] Verbal irony Sperber and Wilson’s theory of verbal irony has been under development from the late 1970s, but has not changed in its essentials since the presentation in Relevance (Sperber & Wilson, 1986, pp. 237– 243), which makes small but significant changes to the original proposal (Sperber & Wilson, 1981). Other publications by Wilson on irony include (Wilson & Sperber, 1992; Sperber & Wilson, 1998a; Wilson, 2006; Wilson, 2009; Wilson & Sperber, 2012a; Wilson, 2013). In verbal irony, a speaker typically says something that she does not endorse herself. In the classical analysis of rhetoric going back to Aristotle – and in dictionaries to this day – verbal irony is defined as “the expression of one’s meaning by using language that normally signifies the opposite” (Merriam Webster). In his seminal work on conversation, the philosopher Paul Grice also adopted this definition, but with misgivings, since he knew of apparent counterexamples such as (1), which falls flat if uttered pointing at a car with a broken window. (1) Look, that car has all its windows intact. (Grice, 1967/1989, p. 53) It is also problematic for the classical definition that a speaker can say something ironically that she herself takes to be true, as in (2) and (3): (2) It seems to be raining.